The increase of smoking prevalence with age occurred among both medical and non-medical students at very similar rates. To facilitate statistical analyses, we grouped the students into three age groups, ≤19, 20–21 and ≥22 years, because some of the age subgroups had small numbers of smokers. The rate of increase from youngest (≤19 years) to oldest (≥22 years) was 59% among medical students and 60% among non-medical students.
Percentages of occasional smokers, which also changed with age, showed a different pattern for medical students. Percentages of occasional smokers among medical students increased with age, from 75% for the ≤19 and 20–21 year groups to 77% for the ≥22 year group. Percentages for non-medical students decreased with age, from 70% for the ≤19 year group to 63% for the 20–21 year and 47% for ≥22 year groups. A test of interaction between age and student group for these percentages showed marginal significance (p = 0.089).
Because prevalence varied with age, the prevalence for medical students was adjusted by age according to the age distribution of non-medical students [although the two groups actually had similar means (SDs): 20.6 (1.5) for medical students, 20.5 (1.6) for non-medical students]. When adjusted by age, the smoking prevalence for medical students dropped slightly from 26.8% to 26.3%, which was not significantly different from the non-medical students' rate of 29.8% (p = 0.10).